Inspiration

You might think that throwing away trash into different categorized bins seems self-explanatory, even trivial, when doing so. But is it really? Try Garbage Sort.

What it does

Garbage Sort provides a webapp to easily categorize and waste material. It detects any object in the webcam and sorts the object into a category of trash, recycle, and compost.

The webapp also can test one's knowledge of different waste objects. In an included game, a user guesses the type of waste in the webcam and is given feedback based on the answer given.

How we built it

For the live camera implementation, we used OpenCV to capture the camera frame by frame. Then the video was analyzed by Google’s Cloud Vision object detection using Machine Learning and returns the annotations of the image. Given certain keyword classifications, our program sorts the Garbage into 3 categories: Trash, Recycle, Compost.

We used Flask to help us in the implementation of the backend, and React to build the front end web page.

Challenges we ran into

We had trouble implementing and setting up relatively, to us, new technologies such as the Google Cloud API. Furthermore, calling data from Flask into React was definitely a learning process.

Accomplishments that we're proud of

We’re proud of the fact that we created a fully functioning app that uses both ends of a web app, something that we’ve never done before.

What we learned

We learned a lot throughout this hackathon. Frameworks, such as Flask and React, were new to us, and we haven’t used any Google Cloud products before.

What's next for Garbage Classification Sort

We hope that we can create more accurate models of waste items to broaden the scope of what the objects can detect.

Built With

Share this project:

Updates